Metraipolis › archive › dc_e6d37ea1d61e
Once Generated, Ranked: End-to-End Generative Slate Recommendation with Unified Semantic-Collaborative IDs — arXiv:2608.17613
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.17613
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T15:23:10.644Z
- SHA-256 of the archived text
421d195a150ea4e45bd3691f2a5fe69305142b1bc938942dc401b7cd081e604e- Stable URL of this page
https://metraipolis.com/documento/dc_e6d37ea1d61e
How to cite
Once Generated, Ranked: End-to-End Generative Slate Recommendation with Unified Semantic-Collaborative IDs — arXiv:2608.17613 — arXiv (en), https://arxiv.org/abs/2608.17613 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e6d37ea1d61e (Metraipolis archive dc_e6d37ea1d61e, sha256 421d195a150ea4e45bd3691f2a5fe69305142b1bc938942dc401b7cd081e604e)
BibTeX
@misc{dce6d37ea1d61e,
title = {Once Generated, Ranked: End-to-End Generative Slate Recommendation with Unified Semantic-Collaborative IDs — arXiv:2608.17613},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.17613},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e6d37ea1d61e; sha256 421d195a150ea4e45bd3691f2a5fe69305142b1bc938942dc401b7cd081e604e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.17613. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Slate recommendation treats a slate rather than an individual item as the recommendation unit, requiring joint optimization of item interactions and slate utility. Existing approaches typically separate candidate generation from ranking and restrict optimization to retrieved candidates. Generative recommendation with Semantic IDs (SIDs) offers a path to end-to-end recommendation, but existing SID construction often lacks recommendation-aware semantics and effective local collaborative signals, while next-token prediction is misaligned with slate-level objectives. We propose OGR, an end-to-end framework that directly generates ordered slates-"Once Generated, Ranked." OGR first introduces TUSID, which adaptively fuses item-specific semantic and local collaborative information into hierarchical SIDs. It then uses list-wise preference planning and pipelined position-wise SID decoding to model global preferences and inter-item dependencies while generating ordered slates. We further propose SPA, a reward-guided conservative policy optimization method that aligns generated slates with user preferences beyond likelihood imitation. Offline experiments show that OGR outperforms representative baselines, with 48.2% and 27.2% relative NDCG@5 gains on industrial and public datasets, respectively. Online A/B testing on Kuaishou further yields a 1.120% improvement in Effective Views. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-19 in cs.IR). Categorie arXiv: Information Retrieval (cs.IR) ; Social and Information Networks (cs.SI). Pagina del lavoro: https://arxiv.org/abs/2608.17613 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-18T15:18:13.755Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-08-19?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).